Contactless monitoring of blood pressure using photoplethysmography

From ISLAB/CAISR
Revision as of 15:54, 18 October 2018 by Taha (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Title Contactless monitoring of blood pressure using photoplethysmography
Summary A camera-based system for non-invasive monitoring of blood pressure
Keywords Photoplethysmography, blood pressure, computer vision, neural networks
TimeFrame Fall 2018
References
Prerequisites Computer vision or digital image processing, digital signal processing, Machine learning, Artificial Intelligence
Author
Supervisor Taha Khan
Level Master
Status Open


In emergency medicine and urgent care, the first assessment and triage are crucial to examine the severity of disease, condition or injury, for prioritizing tests and procedures. Vital signs are fundamental in this process and are investigated in all validated clinical assessment methods. This proposal aims to develop a camera-based system based on photoplethysmography to enable non-contact scanning of vital signs, specifically pulse rate and blood pressure (BP), replacing the conventional setup of devices that are time-consuming and require physical contact. The method will work by recording a video for a few seconds using a high-speed RGB camera. This video will then be processed by image and signal processing algorithms, and machine learning to classify blood pressure, and trigger a flag in case if a patient is in need of emergency care, enabling the staff in the care unit to decide on the level of caution to be taken.

Research Question: How to classify blood pressure levels using photoplethysmography and machine learning?